When data privacy is imposed as a necessity, Federated learning (FL) emerges as a relevant artificial intelligence field for developing machine learning (ML) models in a distributed and decentralized environment. FL allows ML models to be trained on local devices without any need for centralized data transfer, thereby reducing both the exposure of sensitive data and the possibility of data interception by malicious third parties. This paradigm has gained momentum in the last few years, spurred by the plethora of real-world applications that have leveraged its ability to improve the efficiency of distributed learning and to accommodate numerous participants with their data sources. By virtue of FL, models can be learned from all such distributed data sources while preserving data privacy. The aim of this paper is to provide a practical tutorial on FL, including a short methodology and a systematic analysis of existing software frameworks. Furthermore, our tutorial provides exemplary cases of study from three complementary perspectives: i) Foundations of FL, describing the main components of FL, from key elements to FL categories; ii) Implementation guidelines and exemplary cases of study, by systematically examining the functionalities provided by existing software frameworks for FL deployment, devising a methodology to design a FL scenario, and providing exemplary cases of study with source code for different ML approaches; and iii) Trends, shortly reviewing a non-exhaustive list of research directions that are under active investigation in the current FL landscape. The ultimate purpose of this work is to establish itself as a referential work for researchers, developers, and data scientists willing to explore the capabilities of FL in practical applications.

Cyber-physical systems (CPSs) have emerged as an essential area of research in the last decade, providing a new paradigm for the integration of computational and physical units in modern control systems. Remote state estimation (RSE) is an indispensable functional module of CPSs. Recently, it has been demonstrated that malicious agents can manipulate data packets transmitted through unreliable channels of RSE, leading to severe estimation performance degradation. This paper aims to present an overview of recent advances in cyber-attacks and defensive countermeasures, with a specific focus on integrity attacks against RSE. Firstly, two representative frameworks for the synthesis of optimal deception attacks with various performance metrics and stealthiness constraints are discussed, which provide a deeper insight into the vulnerabilities of RSE. Secondly, a detailed review of typical attack detection and resilient estimation algorithms is included, illustrating the latest defensive measures safeguarding RSE from adversaries. Thirdly, some prevalent attacks impairing the confidentiality and data availability of RSE are examined from both attackers’ and defenders’ perspectives. Finally, several challenges and open problems are presented to inspire further exploration and future research in this field.

This paper proposes linear and nonlinear filters for a non-Gaussian dynamic system with an unknown nominal covariance of the output noise. The challenge of designing a suitable filter in the presence of an unknown covariance matrix is addressed by focusing on the output data set of the system. Considering that data generated from a Gaussian distribution exhibit ellipsoidal scattering, we first propose the weighted sum of norms (SON) clustering method that prioritizes nearby points, reduces distant point influence, and lowers computational cost. Then, by introducing the weighted maximum likelihood, we propose a semi-definite program (SDP) to detect outliers and reduce their impacts on each cluster. Detecting these weights paves the way to obtain an appropriate covariance of the output noise. Next, two filtering approaches are presented: a cluster-based robust linear filter using the maximum a posterior (MAP) estimation and a cluster-based robust nonlinear filter assuming that output noise distribution stems from some Gaussian noise resources according to the ellipsoidal clusters. At last, simulation results demonstrate the effectiveness of our proposed filtering approaches.

This paper is concerned with consensus of a second-order linear time-invariant multi-agent system in the situation that there exists a communication delay among the agents in the network. A proportional-integral consensus protocol is designed by using delayed and memorized state information. Under the proportional-integral consensus protocol, the consensus problem of the multi-agent system is transformed into the problem of asymptotic stability of the corresponding linear time-invariant time-delay system. Note that the location of the eigenvalues of the corresponding characteristic function of the linear time-invariant time-delay system not only determines the stability of the system, but also plays a critical role in the dynamic performance of the system. In this paper, based on recent results on the distribution of roots of quasi-polynomials, several necessary conditions for Hurwitz stability for a class of quasi-polynomials are first derived. Then allowable regions of consensus protocol parameters are estimated. Some necessary and sufficient conditions for determining effective protocol parameters are provided. The designed protocol can achieve consensus and improve the dynamic performance of the second-order multi-agent system. Moreover, the effects of delays on consensus of systems of harmonic oscillators/double integrators under proportional-integral consensus protocols are investigated. Furthermore, some results on proportional-integral consensus are derived for a class of high-order linear time-invariant multi-agent systems.

Blast furnace (BF) burden surface contains the most abundant, intuitive and credible smelting information and acquiring high-definition and high-brightness optical images of which is essential to realize precise material charging control, optimize gas flow distribution and improve ironmaking efficiency. It has been challengeable to obtain high-quality optical burden surface images under high-temperature, high-dust, and extremely-dim (less than 0.001 Lux) environment. Based on a novel endoscopic sensing detection idea, a reverse telephoto structure starlight imaging system with large field of view and large aperture is designed. Combined with a water-air dual cooling intelligent self-maintenance protection device and the imaging system, a starlight high-temperature industrial endoscope is developed to obtain clear optical burden surface images stably under the harsh environment. Based on an endoscope imaging area model, a material flow trajectory model and a gas-dust coupling distribution model, an optimal installation position and posture configuration method for the endoscope is proposed, which maximizes the effective imaging area and ensures large-area, safe and stable imaging of the device in a confined space. Industrial experiments and applications indicate that the proposed method obtains clear and reliable large-area optical burden surface images and reveals new BF conditions, providing key data support for green iron smelting.

This paper investigates the tracking control problem for unmanned underwater vehicles (UUVs) systems with sensor faults, input saturation, and external disturbance caused by waves and ocean currents. An active sensor fault-tolerant control scheme is proposed. First, the developed method only requires the inertia matrix of the UUV, without other dynamic information, and can handle both additive and multiplicative sensor faults. Subsequently, an adaptive fault-tolerant controller is designed to achieve asymptotic tracking control of the UUV by employing robust integral of the sign of error feedback method. It is shown that the effect of sensor faults is online estimated and compensated by an adaptive estimator. With the proposed controller, the tracking error and estimation error can asymptotically converge to zero. Finally, simulation results are performed to demonstrate the effectiveness of the proposed method.

Solving constrained multi-objective optimization problems with evolutionary algorithms has attracted considerable attention. Various constrained multi-objective optimization evolutionary algorithms (CMOEAs) have been developed with the use of different algorithmic strategies, evolutionary operators, and constraint-handling techniques. The performance of CMOEAs may be heavily dependent on the operators used, however, it is usually difficult to select suitable operators for the problem at hand. Hence, improving operator selection is promising and necessary for CMOEAs. This work proposes an online operator selection framework assisted by Deep Reinforcement Learning. The dynamics of the population, including convergence, diversity, and feasibility, are regarded as the state; the candidate operators are considered as actions; and the improvement of the population state is treated as the reward. By using a Q-network to learn a policy to estimate the Q-values of all actions, the proposed approach can adaptively select an operator that maximizes the improvement of the population according to the current state and thereby improve the algorithmic performance. The framework is embedded into four popular CMOEAs and assessed on 42 benchmark problems. The experimental results reveal that the proposed Deep Reinforcement Learning-assisted operator selection significantly improves the performance of these CMOEAs and the resulting algorithm obtains better versatility compared to nine state-of-the-art CMOEAs.

The success of deep transfer learning in fault diagnosis is attributed to the collection of high-quality labeled data from the source domain. However, in engineering scenarios, achieving such high-quality label annotation is difficult and expensive. The incorrect label annotation produces two negative effects: 1) the complex decision boundary of diagnosis models lowers the generalization performance on the target domain, and 2) the distribution of target domain samples becomes misaligned with the false-labeled samples. To overcome these negative effects, this article proposes a solution called the label recovery and trajectory designable network (LRTDN). LRTDN consists of three parts. First, a residual network with dual classifiers is to learn features from cross-domain samples. Second, an annotation check module is constructed to generate a label anomaly indicator that could modify the abnormal labels of false-labeled samples in the source domain. With the training of relabeled samples, the complexity of diagnosis model is reduced via semi-supervised learning. Third, the adaptation trajectories are designed for sample distributions across domains. This ensures that the target domain samples are only adapted with the pure-labeled samples. The LRTDN is verified by two case studies, in which the diagnosis knowledge of bearings is transferred across different working conditions as well as different yet related machines. The results show that LRTDN offers a high diagnosis accuracy even in the presence of incorrect annotation.

This paper proposes a novel event-driven encrypted control framework for linear networked control systems (NCSs), which relies on two modified uniform quantization policies, the Paillier cryptosystem, and an event-triggered strategy. Due to the fact that only integers can work in the Pailler cryptosystem, both the real-valued control gain and system state need to be first quantized before encryption. This is dramatically different from the existing quantized control methods, where only the quantization of a single value, e.g., the control input or the system state, is considered. To handle this issue, static and dynamic quantization policies are presented, which achieve the desired integer conversions and guarantee asymptotic convergence of the quantized system state to the equilibrium. Then, the quantized system state is encrypted and sent to the controller when the triggering condition, specified by a state-based event-triggered strategy, is satisfied. By doing so, not only the security and confidentiality of data transmitted over the communication network are protected, but also the ciphertext expansion phenomenon can be relieved. Additionally, by tactfully designing the quantization sensitivities and triggering error, the proposed event-driven encrypted control framework ensures the asymptotic stability of the overall closed-loop system. Finally, a simulation example of the secure motion control for an inverted pendulum cart system is presented to evaluate the effectiveness of the theoretical results.

This paper investigates the path-following control problem with obstacle avoidance of autonomous surface vehicles in the presence of actuator faults, uncertainty and external disturbances. Autonomous surface vehicles inevitably suffer from actuator faults in complex sea environments, which may cause existing obstacle avoidance strategies to fail. To reduce the influence of actuator faults, an improved artificial potential function is constructed by introducing the lower bound of actuator efficiency factors. The nonlinear state observer, which only depends on measurable position information of the autonomous surface vehicle, is used to address uncertainties and external disturbances. By using a backstepping technique and adaptive mechanism, a path-following control strategy with obstacle avoidance and fault tolerance is designed which can ensure that the tracking errors converge to a small neighborhood of zero. Compared with existing results, the proposed control strategy has the capability of obstacle avoidance and fault tolerance simultaneously. Finally, the comparison results through simulations are given to verify the effectiveness of the proposed method.

In a crowd density estimation dataset, the annotation of crowd locations is an extremely laborious task, and they are not taken into the evaluation metrics. In this paper, we aim to reduce the annotation cost of crowd datasets, and propose a crowd density estimation method based on weakly-supervised learning, in the absence of crowd position supervision information, which directly reduces the number of crowds by using the number of pedestrians in the image as the supervised information. For this purpose, we design a new training method, which exploits the correlation between global and local image features by incremental learning to train the network. Specifically, we design a parent-child network (PC-Net) focusing on the global and local image respectively, and propose a linear feature calibration structure to train the PC-Net simultaneously, and the child network learns feature transfer factors and feature bias weights, and uses the transfer factors and bias weights to linearly feature calibrate the features extracted from the Parent network, to improve the convergence of the network by using local features hidden in the crowd images. In addition, we use the pyramid vision transformer as the backbone of the PC-Net to extract crowd features at different levels, and design a global-local feature loss function (L₂). We combine it with a crowd counting loss (L_C) to enhance the sensitivity of the network to crowd features during the training process, which effectively improves the accuracy of crowd density estimation. The experimental results show that the PC-Net significantly reduces the gap between fully-supervised and weakly-supervised crowd density estimation, and outperforms the comparison methods on five datasets of ShanghaiTech Part A, ShanghaiTech Part B, UCF_CC_50, UCF_QNRF and JHU-CROWD++.

Oscillation detection has been a hot research topic in industries due to the high incidence of oscillation loops and their negative impact on plant profitability. Although numerous automatic detection techniques have been proposed, most of them can only address part of the practical difficulties. An oscillation is heuristically defined as a visually apparent periodic variation. However, manual visual inspection is labor-intensive and prone to missed detection. Convolutional neural networks (CNNs), inspired by animal visual systems, have been raised with powerful feature extraction capabilities. In this work, an exploration of the typical CNN models for visual oscillation detection is performed. Specifically, we tested MobileNet-V1, ShuffleNet-V2, EfficientNet-B0, and GhostNet models, and found that such a visual framework is well-suited for oscillation detection. The feasibility and validity of this framework are verified utilizing extensive numerical and industrial cases. Compared with state-of-the-art oscillation detectors, the suggested framework is more straightforward and more robust to noise and mean-nonstationarity. In addition, this framework generalizes well and is capable of handling features that are not present in the training data, such as multiple oscillations and outliers.

This paper develops a quadratic function convex approximation approach to deal with the negative definite problem of the quadratic function induced by stability analysis of linear systems with time-varying delays. By introducing two adjustable parameters and two free variables, a novel convex function greater than or equal to the quadratic function is constructed, regardless of the sign of the coefficient in the quadratic term. The developed lemma can also be degenerated into the existing quadratic function negative-determination (QFND) lemma and relaxed QFND lemma respectively, by setting two adjustable parameters and two free variables as some particular values. Moreover, for a linear system with time-varying delays, a relaxed stability criterion is established via our developed lemma, together with the quivalent reciprocal combination technique and the Bessel-Legendre inequality. As a result, the conservatism can be reduced via the proposed approach in the context of constructing Lyapunov-Krasovskii functionals for the stability analysis of linear time-varying delay systems. Finally, the superiority of our results is illustrated through three numerical examples.

The trajectory tracking control performance of nonholonomic wheeled mobile robots (NWMRs) is subject to nonholonomic constraints, system uncertainties, and external disturbances. This paper proposes a barrier function-based adaptive sliding mode control (BFASMC) method to provide high-precision, fast-response performance and robustness for NWMRs. Compared with the conventional adaptive sliding mode control, the proposed control strategy can guarantee that the sliding mode variables converge to a predefined neighborhood of origin with a predefined reaching time independent of the prior knowledge of the uncertainties and disturbances bounds. Another advantage of the proposed algorithm is that the control gains can be adaptively adjusted to follow the disturbances amplitudes thanks to the barrier function. The benefit is that the overestimation of control gain can be eliminated, resulting in chattering reduction. Moreover, a modified barrier function-like control gain is employed to prevent the input saturation problem due to the physical limit of the actuator. The stability analysis and comparative experiments demonstrate that the proposed BFASMC can ensure the pre-specified convergence performance of the NWMR system output variables and strong robustness against uncertainties/disturbances.

Powered by advanced information technology, more and more complex systems are exhibiting characteristics of the cyber-physical-social systems (CPSS). In this context, computational experiments method has emerged as a novel approach for the design, analysis, management, control, and integration of CPSS, which can realize the causal analysis of complex systems by means of “algorithmization” of “counterfactuals”. However, because CPSS involve human and social factors (e.g., autonomy, initiative, and sociality), it is difficult for traditional design of experiment (DOE) methods to achieve the generative explanation of system emergence. To address this challenge, this paper proposes an integrated approach to the design of computational experiments, incorporating three key modules: 1) Descriptive module: Determining the influencing factors and response variables of the system by means of the modeling of an artificial society; 2) Interpretative module: Selecting factorial experimental design solution to identify the relationship between influencing factors and macro phenomena; 3) Predictive module: Building a meta-model that is equivalent to artificial society to explore its operating laws. Finally, a case study of crowd-sourcing platforms is presented to illustrate the application process and effectiveness of the proposed approach, which can reveal the social impact of algorithmic behavior on “rider race”.